All in good time

When it comes to vessel autonomy, Damen’s Principal Research Engineer, Automation, Dimitrios Kotiadis explains, it’s a case of evolution, not revolution

Damen Shipyards Group continues to invest in research into autonomous shipping. The company has previously released both an article and paper covering its activities in this field and detailing its See – Think – Act approach.

In a nutshell, this covers the idea that, to replace human functionality, a vessel needs integrated technology to observe what is taking place around it, analyse the information, and make sound judgements on how to respond to what it sees.

Dimitrios Kotiadis

Tooled up

To achieve this capability, says Dimitrios Kotiadis, Principal Research Engineer, Automation at Damen, requires a variety of tools. Among these is a simulator. Damen has approached this head-on, constructing its own purpose-built simulation tool.

“We had to,” Dimitrios says. “They are not so widely available, and when you can source one, chances are it’s been designed to facilitate human, not machine training.”

Simulation is a vital part of the path to automation, he explains, providing the opportunity to test things, again and again, under the exact same conditions.

“You cannot do that in the real world. No matter how hard you try, you cannot replicate the conditions precisely. With a simulation, you can have a process of test, adjust, test again, tweaking things until they work as you wish them to.”

A step into reality

Be that as it may, there are limitations, and simulation represents only a piece of the puzzle.

“As essential as simulation is to the process, and no matter how realistic it looks, something is missing. You are never going to get the full picture. You won’t get the sensor effect, or the weather conditions, for example.”

For this, Dimitrios continues, you need to take a step into the real world. For Damen, this means DAVe – the Damen Autonomous Vessel, a model ship approximately 1.1 metres in length that Damen has extensively tested, first of all in a swimming pool, and later in the harbour at its yard in Gorinchem on the River Merwede.

“With a model like DAVe, you are unable to repeat everything, of course. What you do have, however, is hydrodynamics – the vessel’s interaction with the water and of course the real world sensor effects. Now it starts to become very different from a simulation and much closer to reality.”

Towards a full-scale solution

DAVe, then, represents an important progression. The model, though, is not without its own limitations.

“With a model, you still have some restrictions. Simply put, it’s very small. We cannot pitch it against the weather, or test how it responds when a larger vessel comes close.”

From here, the way forward is to go full-scale. This is where Damen finds itself at the time of writing. The company has repurchased a Damen-built workboat it delivered some years ago, and is equipping it to See – Think – Act.

“This is an important phase,” states Dimitrios. “It places us firmly in the real world. With this, we can sail at night, or during harsh weather, for example. We can start to harvest real data.”

This, he says, paves the way for the fourth – and final – step: deployment. With that stage, a continual stream of data is produced, showing, for instance, what is happening during an operation, how the systems are behaving, when something fails, and why.

Towards a full-scale solution

DAVe, then, represents an important progression. The model, though, is not without its own limitations.

“With a model, you still have some restrictions. Simply put, it’s very small. We cannot pitch it against the weather, or test how it responds when a larger vessel comes close.”

From here, the way forward is to go full-scale. This is where Damen finds itself at the time of writing. The company has repurchased a Damen-built workboat it delivered some years ago, and is equipping it to See – Think – Act.

“This is an important phase,” states Dimitrios. “It places us firmly in the real world. With this, we can sail at night, or during harsh weather, for example. We can start to harvest real data.”

This, he says, paves the way for the fourth – and final – step: deployment. With that stage, a continual stream of data is produced, showing, for instance, what is happening during an operation, how the systems are behaving, when something fails, and why.